1583565 - Crash in [@ java.lang.NullPointerException: at org.mozilla.gecko.media.GeckoHlsPlayer.getVideoInfo(GeckoHlsPlayer.java)]

Status: RESOLVED FIXED

(Core :: Audio/Video: Playback, defect, P2)

Description

[Marcia Knous [:marcia]]

This bug is for crash report bp-1f49cad4-9275-4238-8f0b-eeac60190924.

Seen while looking at beta specific crashes - not a new regression but goes back to at least 68.0. Many mentions of crashing while browsing the reddit mobile site: https://mzl.la/2mulNoj.

here are some URLs:

• https://homeguides.sfgate.com/remove-residue-bottom-steam-iron-34138.html
• https://livforcake.com/smores-nachos/
• https://www.bowlofdelicious.com/instant-pot-split-pea-soup-ham-bone/
• https://www.livenewsnow.com/american/msnbc.html
• https://www.newsweek.com/my-hero-academia-season-4-release-date-1267349
• https://www.reddit.com/r/Archero/comments/btxw3v/new_ability_tier_list/

Various flavors of Pixel running API 29 are in the mix. Perhaps we could try to reproduce this crash?

Java stack trace:

java.lang.NullPointerException
	at org.mozilla.gecko.media.GeckoHlsPlayer.getVideoInfo(GeckoHlsPlayer.java:717)
	at org.mozilla.gecko.media.GeckoHLSDemuxerWrapper.getVideoInfo(GeckoHLSDemuxerWrapper.java:104)

Comment 1

[Stefan Arentz | :st3fan | ⏰ EST | he/him]

Please move back to Fennec if GeckoView is not the proper product/component for this bug.

Comment 2

[Marcia Knous [:marcia]]

There are some crashes in Focus, but the lion's share of crashes are Fennec. The Focus crashes are happening on Google Pixel devices as well.

Comment 3

[Stefan Deiac]

Hello, I tried to reproduce this crash using the information from comment 0 but I wasn't able.

Devices used

• Samsung Galaxy S8 (Android 9);
• Google Pixel 2 (Android 9).

Build:

• Firefox 68.1.1 - playstore.

If there is any information on this crash or additional help about this ticket, please let me know and I will help, thanks.

Comment 4

[Nils Ohlmeier [:drno]]

Just from looking at mercurial logs I think this bug 1552145 is a highly likely candidate for causing these problems.

Comment 5

[Chris Peterson [:cpeterson]]

This crash signature really spiked with the release of Fennec ESR 68.1 (September 3), but started ramping up in July 9 (either with the release of Fennec ESR 68.0 or the 68.1b beta builds in the Fennec Nightly and Beta channels).

Comment 6

[Julien Cristau [:jcristau]]

[Tracking Requested - why for this release]: fennec crash regression

Comment 7

[Paul Adenot (:padenot)]

John, how familiar are you with the HLS player code ?

Comment 8

[Ryan VanderMeulen [:RyanVM]]

Nils, this is still pretty high volume. Is there anybody with cycles to look into it?

Comment 9

[John Lin [:jhlin][:jolin]]

(In reply to Ryan VanderMeulen [:RyanVM] from comment #8)

Nils, this is still pretty high volume. Is there anybody with cycles to look into it?

I'll check it.

Comment 10

[BugBot [:suhaib / :marco/ :calixte]]

The priority flag is not set for this bug.
:drno, could you have a look please?

For more information, please visit auto_nag documentation.

Comment 11

[Liz Henry (:lizzard) (relman/hg->git project)]

Is this still something you might investigate? It's still high volume for Fennec and present in lower volume in Focus as well.

Comment 12

[John Lin [:jhlin][:jolin]]

(In reply to Liz Henry (:lizzard) from comment #11)

Is this still something you might investigate? It's still high volume for Fennec and present in lower volume in Focus as well.

Sorry for not updating the status for so long. Although unable to reproduce it locally, I did some code study and thought the symptom could be mitigated. I'll try to land the patch after completing bug 1592140.

Comment 13

[John Lin [:jhlin][:jolin]]

Attachment: Bug 1583565 - check if Java wrapper is still valid when updating media info. r?jya

Comment 14

[John Lin [:jhlin][:jolin]]

Attachment: Bug 1583565 - check if Java wrapper is still valid when updating media info. r?jya

Comment 15

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/82fd8c769ea9
check if Java wrapper is still valid when updating media info. r=jya

Comment 16

[Cristian Brindusan [:cbrindusan]]

https://hg.mozilla.org/mozilla-central/rev/82fd8c769ea9

Comment 17

[Ryan VanderMeulen [:RyanVM]]

Please nominate this for ESR68 approval when you get a chance. It grafts cleanly as-landed.

Comment 18

[John Lin [:jhlin][:jolin]]

Comment on attachment 9115545 [details]
Bug 1583565 - check if Java wrapper is still valid when updating media info. r?jya

ESR Uplift Approval Request

• If this is not a sec:{high,crit} bug, please state case for ESR consideration: This patch fixes crashes that were reported thousands of times.
• User impact if declined: Occasional crashes when playing HLS contents.
• Fix Landed on Version: 73
• Risk to taking this patch: Low
• Why is the change risky/not risky? (and alternatives if risky): It is just a minor change that checks for null pointers.
• String or UUID changes made by this patch:

Comment 19

[Ryan VanderMeulen [:RyanVM]]

Comment on attachment 9115545 [details]
Bug 1583565 - check if Java wrapper is still valid when updating media info. r?jya

Fixes a frequent Fennec crash. Approved for 68.5b2.

Comment 20

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-esr68/rev/b783e1f8ce98

Comment 21

[Diana Rus]

Hi, i tried to reproduce the crash with Google Pixel 3a XL(Android 10) -api level 29 on an aarch64 build for version 68.3.0, but i was not able to reproduce it with the provided info from description and neither with info from comment 4. I also do not reproduce the crash on 68.5b2 on the same device on an aarch64.

We will wait a few days until marking as verified to see whether after someone reproduces it.

Comment 22

[Ryan VanderMeulen [:RyanVM]]

Yeah, I think we can just keep an eye on crash-stats for a bit to verify that the crash is gone.

Comment 23

[Chris Peterson [:cpeterson]]

Removing regressionwindow-wanted keyword because this bug has been resolved.

Comment 24

[Chris Peterson [:cpeterson]]

Removing regressionwindow-wanted keyword because this bug has been resolved.

Comment 25

[Chris Peterson [:cpeterson]]

Removing regressionwindow-wanted keyword because this bug has been resolved.